Geoffrey Siwo: Crowdsourcing Cures
WHY YOU SHOULD CARE
Geoffrey Siwo wants to accelerate drug discovery by enlisting the help of millions of Internet users around the world — not research insitutions or Big Pharma.
By Melissa Pandika
There are more than 10,000 diseases on the planet, asthma to zygomycosis. Geoffrey Siwo wants to cure all of them.
And the 36-year-old computational biologist believes he can. Or, rather, that we can, if we can disrupt drug discovery. It’s worth listening: Few understand the toll of disease like Siwo, who lost three sisters to disease in Kenya.
Today, pharmaceutical companies pour resources into widespread diseases that afflict developed countries, like cancer or diabetes. There’s less economic incentive to tackle rare diseases or those, like malaria and tuberculosis, that plague poorer regions. Even if there were, drug companies would still lack the manpower to tackle every disease known.
What if you could find a way in which basically anyone with a computer and Internet connection could contribute to research on a disease?
— Geoffrey Siwo
Siwo sees another way. “What if you could find a way in which basically anyone with a computer and Internet connection could contribute to research on a disease?” he asks — and it’s not a rhetorical question. This month, he unveiled the United Genomes Project, Africa’s first crowdsourced, open-source genetic database. Siwo and his team will start by asking first-generation African-Americans to upload their results from 23andMe and other commercially available genetic kits. Then they’ll fan out to African countries once they approve direct-to-consumer genetic testing.
Genomic data could help shed light on mutations and other factors that underlie diseases and susceptibility to them, as well as drug responses, allowing physicians to tailor treatments to individual patients. United Genomes’ database would also teach students to analyze genetic information, allowing even those without lab access to generate and answer their own research questions.
In 2011, Foldit players deciphered the structure of an AIDS-causing monkey virus in 10 days — a puzzle that had stumped scientists for 15 years.
Thirty-five-year-old Siwo has velvety eyes and a schoolboy expression that breaks readily into a smile. He speaks softly but openly from his office at Dartmouth, where he’s a postdoc, describing his childhood in dusty, rural Homa Bay, Kenya — sans electricity or running water. One sister died of malaria when she was 9, and it shaped him.
As did his teachers. A middle school teacher kindled his passion for biology when she described cancer as uncontrolled cell growth, spurring him to read voraciously about how the disease developed. At 19, Siwo wrote to the then-chair of Stanford University’s biology department, Patricia Jones, asking what determined the arrangement of genes in DNA. Six months later, he received a parcel from her: a college-level molecular and cell biology textbook. He devoured all 1,000-plus pages.
As an undergraduate at Egerton University, in Nairobi, Siwo conducted research at the Institute of Primate Research. Tucked in the Oloolua forest, where giant trees rustle with baboons, Siwo studied endogenous retroviruses. Retroviruses store genetic information in the form of a molecule called RNA. When they infect a host, an enzyme called reverse transcriptase converts RNA is converted to DNA, which then inserts itself into the host’s genome. Endogenous retroviruses are fragments of retroviral DNA from past infections, inherited from earlier generations.
Knowing that HIV is a retrovirus, too, Siwo wondered, “How could these two retroviruses interact?” He hypothesized that HIV could borrow genes from endogenous retroviruses to resist HIV drugs. Our genomes have accumulated so many different retroviral DNA sequences that chance suggests that at least some have mutations that confer drug resistance.
There was doubt about whether Siwo could test his hypothesis without large-scale molecular experiments, which would require large-scale funding. So he ditched traditional molecular methods: “I thought, ’What if I could test it on a computer?’”
That … convinced me of the power of computing as a means of solving very difficult biological problems.
— Geoffrey Siwo
Siwo didn’t own a computer; he’d barely even used one. So he clocked in an hour or two a day at a local cybercafé, six days a week for several months, analyzing human DNA sequences available on an online database. The café charged him a dollar an hour — more than he had spent on food in a day. Siwo often skipped lunches to save money.
Less than 10 percent of available genetic data come from Africans, Siwo says. Building out a database could lead to better information and better medicine.
Sure enough, he observed that endogenous retrovirus reverse transcriptase contains mutations that can confer resistance to HIV drugs. That means if HIV incorporates some of these mutated sequences into its own DNA, the result could be a drug-resistant virus. He submitted an abstract to the Interscience Conference on Antimicrobial Agents and Chemotherapy — and, at 24, received an invitation to present his findings in Chicago. “That experience has always convinced me of the power of computing as a means of solving very difficult biological problems,” he says. He hopes the United Genomes Project database could similarly help other African scientists conduct their own research.
Siwo then headed to the University of Notre Dame for a doctorate in computational biology, investigating drug resistance and malaria. His Ph.D. adviser, Gustavo Stolovitzky, recalls him as shy yet persistent. “If he has a question, he forces himself to ask it. It’s more important for him to know,” he says. “He’s a gem, but a rough gem, polishing himself, working actively on his own development.”
Shyness is a comparatively minor hurdle. In 2007, his two remaining sisters died of pneumonia within the span of a year. “We had grown so close,” he says with a pause. “But their kids are always a reminder. They have an even bigger struggle than I can ever have.”
With United Genomes, Siwo joins the ranks of citizen science advocates like David Baker, who uses games to predict the structure of proteins, and the ILIAD Project, which asks citizen scientists to test plants and other specimens for antibiotic properties. Critics of the citizen science movement question the rigor of data collected by amateurs. But there are clear advantages to crowdsourcing genetic data, especially when it comes to data from African countries. Less than 10 percent of available genetic data come from Africans, Siwo says, and building out a database could lead to better information and better medicine for not only Africans, but all other races.
He’s got good timing. A digital revolution swept Africa in the latter half of the last decade, thanks to the installation of undersea fiber optic cables that boosted data transmission capacity. The number of Internet users in Africa grew a whopping 3,600 percent between 2002 and 2012, to 167 million.
Investors seem drawn to his quiet warmth and depth of knowledge. “He gets what he wants when he meets an influential person,” says biomedical consultant and Siwo’s longtime friend, Susan Morgensztern. “They’re in his camp.” He and his colleagues recently licensed a method to predict age from DNA to an undisclosed company, and they are in talks with another company to further develop the technology.
Siwo is still a bachelor. He doesn’t sleep much, according to Stolovitzky, who worries about Siwo “spreading out too thin,” he says. “At this stage of his career … he could focus a little bit more to get the most of his projects.”
But Siwo remains as dogged as ever. “We need to find new mechanisms that can work across multiple diseases,” he says. “Every disease deserves to be worked on.” When it comes to eradicating disease, he can’t spread himself thin enough.
- Melissa Pandika Contact Melissa Pandika